Selective Data Replication for Storage Read Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional network storage systems face inefficiencies in read access operations due to limited access paths and resource utilization, as they often require duplicating entire volumes to increase access paths, which wastes processing resources and storage space.
Innovation Solution
A system that analyzes metadata from access operations to selectively duplicate frequently accessed data items onto additional read volumes, creating smaller, more efficient read volumes that can be accessed in parallel with the source volume, thereby optimizing resource usage and reducing storage space.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If mirror volumes are created to increase access paths, then read access efficiency is improved, but processing resources and storage space are wasted
Solution Approach 1:
The patent extracts only the frequently accessed data items from the source volume to create additional read volumes, rather than duplicating entire volumes. This selective extraction is achieved by analyzing metadata to identify hot data items and copying only those to additional volumes, thereby reducing storage space consumption while maintaining read access efficiency.
Solution Approach 2:
The patent applies local quality by treating different data items differently based on their access characteristics. Frequently accessed data items are replicated to additional volumes while less frequently accessed items remain only on the source volume. This differentiated approach optimizes read performance for hot data without wasting storage space on cold data replication.
2Productivity
If mirror volumes are created to increase access paths, then read access efficiency is improved, but processing resources are consumed
Solution Approach 1:
The system extracts only necessary data items (frequently accessed ones) for replication rather than processing and copying entire volumes. This selective approach significantly reduces the processing resources required for volume creation and updates, while still achieving the goal of increasing access paths for read operations.
Solution Approach 2:
Instead of performing complete volume replication, the patent applies partial action by replicating only a subset of data items that are frequently accessed. This partial replication approach consumes fewer processing resources while providing sufficient read access improvement for the majority of read operations.
3Adaptability or versatility
If entire volumes are duplicated to create additional access paths, then read access concurrency is improved, but storage space is wasted
Solution Approach 1:
The patent extracts only the essential frequently accessed data items from the source volume to populate additional read volumes. This selective extraction creates multiple access paths for hot data without requiring full volume duplication, thereby improving access path versatility while minimizing storage space consumption.
Solution Approach 2:
The patent segments the source volume data into frequently accessed items and less frequently accessed items based on metadata analysis. Only the frequently accessed segment is replicated to additional volumes, creating targeted access paths where needed while avoiding unnecessary replication of the entire volume.
Data Source
AI summary
A computer program product having a computer readable medium tangibly recording computer program logic for providing data access in a storage system, the computer program product including code to analyze a plurality of data access requests directed to a source volume of the storage system, code to generate a mirror volume of the source volume, the mirror volume including a subset of data items of the source volume, in which the first subset of data items is selected for the minor volume based at least in part on data access patterns revealed by analyzing the data access requests, and code to access the mirror volume in parallel with accessing the source volume to fulfill read requests from a plurality of hosts.